Week 1 : Introduction

인공지능

Week 1 : Introduction

1

강의내용


학문

vs.
기술

vs.
철학적인

문제
들


인공지능의

고전적이고

기본적인

이론들



탐색


지식표현과

추론


학습과

예측

…



IT
기술의

최근

이슈에

대한

조사



향후

변화된

상황에서의

새로운

이슈들과

남겨진

문제들

2

강의

운영

방법


수업

:
이론

및

실습


이론

: AI, Data Mining, Search
등


실습

:


탐색방법

: C
나

Java


지식표현

및

추론

: Lisp + KM(Knowledge Machine)


데이터마이닝

: SPSS, Clementine
등

.




발표

:
주제

및

최근

이슈에

관한

조사

발표



개인

과제

및

팀

프로젝트


팀

멤버는

3~4
명





3

평가방법


반영비율




개인

성취율

:


중간고사
: 30%


기말고사
: 30%


과제

및

팀

프로젝트

: 30%


출석

및

수업참여도

: 10%


(
변경될

수

있음
.)

4

상대평가

구간


최소
최대
A
20
~
35
B
20
~
40
C,D
25
~
50
F
0
~
5


Week 1 ~ 2 : AI Introduction

5

What is Artificial Intelligence?


Study of how to make Computers do things which (at the moment) people
do better



Vague?


Different
View Points


Engineering
, Science, Philosophy

Artificial Intelligence: Definition


McCarthy


‘
the science and engineering
of


making intelligent
machine, especially
intelligent computer programs


‘
using
computers to
understand human intelligence
’

What is Intelligence?


Aspects of
Intelligence


problem solving


memorize logical reasoning , intuition
, judgment,
creativity …


learning
commonsense
..


emotion
, cognition, love, hate,.. etc.



What is intelligence?


we
know it when we see it
(
relative concept
)



What is the most basic intelligence?

Turing Test


Ultimate Intelligence: Turing

suggested


Imitation Game (next slide)


Intelligent
as much

as
Human


Is dog intelligent ?


Any man
-
made system passed Turing Test ?


--

Any Examples in SF Movie?

Imitation Game


Intelligent System

Cognition

Understanding

Judgment

Flexible

Automated

Optimized



AI : Engineering Aspects


Making Computer (or

IT
Systems)
more Intelligent
–

better
performance
(
performance? )



Making Machines more
User Friendly



Making a Thinking Machine :
Robot


Can machine think ?

Examples of AI Systems


Intelligent
Home Appliances


Intelligent Building


HCI (Human Computer Interaction)


Intelligent Traffic Control


Robots


Voice, Character Recognition


Ubiquitous System

Cognitive Science


Artificial
Intelligence


Psychology



Neuro
-
Science


Linguistics



Philosophy

Program(algorithm) = mind?

Mind Model

Is mind a chemical reaction?

In Search of Semantics

Can Machine have a Mind?

Brief History of
AI


’
50


1956
–

Dartmouth Conference


MaCarthy
,
Minsky
, Newell


Lisp


60


GPS(General
Problem Solver)
–

Newell


Chess
Programs


’
70


Theorem
Proving
–

resolution(Robinson)


Prolog



’
80



Expert Systems
–

Mycin
,
Prospector



Neural Net


’
90
-

Present



Software Agent, Data Mining



Semantic Web



Ontology

Success / Failure


Sad Story of Machine Translation


Compiler
: Programming Language


Can
you do the same to human language?


“
time flies like an arrow
”



Bonanza



Prospector : first AI system

of
commercial success



Challenge == Machine Understanding!!

AI Impact


Programming Language


Lisp
, Prolog, Object Oriented Language


Database


Knowledge
-
base
, Ontology, NL Query


Internet


Semantic
web, XML


Network


Ubiquitous, Bio
-
Informatics, etc.

Approaches of AI Systems


Knowledge
-
based Approach

(Top Down, Deduction, Symbolic)

u

䑡瑡t䑲iv敮e䅰灲o慣a

⡂o瑴t洠啰Ⱐ䥮摵捴Io測n乥瑷Nr欩

u


Knowledge
-
based System


Represent
Human knowledge

as symbol combination (Rule)

u


䭮潷汥摧攠䅣煵qsi瑩o渠慮搠剥灲敳敮瑡瑩on

䑥D畣瑩u攠卹s瑥t

u


䱯杩挬c䕸灥r琠卹s瑥洬t䙵空礠䱯杩c


Data Driven Approach


Extract common characteristics from collected examples(data)

u


Tr慩湩湧⡃orr散琯e湣潲r散琠䑡瑡t


u


却S瑩s瑩捡c 䵥M桯hⰠ䅲瑩晩捩fl 乥畲慬 乥瑷Nrk

䑡瑡t䵩湩湧


Generality vs Performance


Trade off


Initial Attempts



䝥湥G慬
偲P扬敭e卯lvi湧n⡆慩l畲u)


Complexity : Toy Problems Only



Recent
AI Systems: Specialized Approach


Knowledge Based Approach


Expert Systems


Machine Translation


Human
vs

AI
Technology


Eye


-

Vision, Character Recognition



Mouth


-

Speech Generation



Arms
Legs


-

Robot Arms, Autonomous Vehicle


-

Intelligent Agent,
Softbot


Brain


-

Knowledge Representation


-

Reasoning/ Planning


-

Machine Learning



Other


-

Natural Language


-

Speech Recognition

Research Areas


Symbolic Programming


Knowledge Representation


Search & Planning


Automated Reasoning


Machine Learning/ Data Mining


Artificial Neural Net


Ontology

AI : Future


Application of AI Technology


Smart Home


Web Auto Translation System


Voice Recognition/ Intelligent
HCI



Unified Paradigm


Symbolic Processing + Neural Processing


Knowledge
-
based + Data Driven

AI : Future

AI in everywhere, AI in nowhere





啢i煵q瑯畳u卹s瑥浳



Softbot

(Software Robot)


Human Computer Interface


Understanding


佮瑯lo杹


(Real) Robot



……


…..


과제


AI
관련

영화

보기



예
) A.I
나

I.
로봇



가능한

다양한

관점으로

감상하여

평하기


공학적


인지과학적


윤리학적


철학적

?


A4 1
장


분량의

소감문

제출


보고서

표지

작성하지

마시고
,
보고서

상단에

학번
/
이름

표기


26